« first day (3534 days earlier)      last day (35 days later) » 

5:20 AM
@PhelypeOleinik It's always the right place, don't listen to people who think LaTeX is an exclusive club.
2 hours later…
6:55 AM
@barbarabeeton I honestly don't know why it was closed in the first place (it's certainly not off-topic...)
1 hour later…
8:00 AM
@Skillmonlikestopanswers.xyz It's re-opened now
@CarLaTeX thanks, I saw it :) And actually the question lead to me finding a bug in expkv-def (those pesky unbalanced \if...\fi)
@Skillmonlikestopanswers.xyz :)
@CarLaTeX (and it'll lead to a new feature)
@Skillmonlikestopanswers.xyz Oooohhh
@CarLaTeX nothing special, just a user-definable undefined-choice handler.
8:05 AM
@Skillmonlikestopanswers.xyz I'm sure it's useful
8:40 AM
@Skillmonlikestopanswers.xyz Notice that @cis wants also that \cmd and \cmd[] behave the same, as well \cmd[map=whatever_the_default_color] behaves the same as \cmd[map] (see chat.stackexchange.com/transcript/message/54835309#54835309) --- we got into several iteration for tex.stackexchange.com/questions/552126/…
@Rmano the last point is just the same as doing a default map = whatever_the_default_color, but thanks for warning me :)
@Rmano the question is, does he still want [map=true] behave as [map] and [map=false] like []...
Before I answer I want to finish the updates to expkv-def, expkv, and expkv-opt (and the one for expkv-def needs a bit of work...)
@Skillmonlikestopanswers.xyz I think that he wants [] and nothing behave the same, and use whatever the default is. [map=false] and [map=true] should act independently from the default, while [map] is always the same as [map=true]. (Writing software specification is hard)
9:07 AM
@Rmano yes, indeed, the reason why we have to update answers so often :)
9:18 AM
@PauloCereda boo!
@yo' ooh :)
The Green Ducks took 4th place (out of 16 teams) yesterday --- not bad!
@yo' not bad at all!
@PauloCereda breakfast
9:33 AM
@DavidCarlisle oh no
@DavidCarlisle was the Loading ... message there from the start or is it new?
@UlrikeFischer I added it on ll a few days ago (after Frank moaned nothing happened when he clicked on latex cgi) so it's new on learnlatex
@UlrikeFischer it's not quite the interaction that I intended (seeing what is happening inside the iframe seems tricky but it always pops up for 1 sec which seems about right time
@DavidCarlisle I just wanted to ask for something like this and was wondering if I had missed it before.
Wait a minute, palomita is popcorn in Spanish?!
@DavidCarlisle Imho at this place it is okay. It needs only a bit more design/spacing. I guess there it could be animated? E.g. add dots over time or show something spinning around?
9:45 AM
But paloma is pigeon!
@UlrikeFischer yes it's just an html div that is removed by javascript in 1 sec anything could be inside that.
@DavidCarlisle What is "Loading" in vietnamese?
@PauloCereda Since you are an expert in Spanish, you may appreciate this that I heard on the radio yesterday: "My four your old son has been learning Spanish during the lockdown, but he hasn't even learnt to say 'Please'"
@PauloCereda I think that is poor for four.
@DavidCarlisle LOL same for Portuguese, I guess. :)
@UlrikeFischer currently it's not in the translation javasvript object so it's Loading ... :-)
9:52 AM
@UlrikeFischer Couldn't we use a image or a Unicode symbol like ⌛ ?
It would save the trouble of translating stuff. :)
@PauloCereda well I did at one point suggest using icons for edit and copy etc for same reason
@DavidCarlisle Great minds think alike
@PauloCereda you could write some texts that are injected at random.
@UlrikeFischer ooh a fortune cookie :)
@UlrikeFischer "update in user and admin mode", "notice the red circle", "texworks is just an editor, you need to install tex",
9:59 AM
@DavidCarlisle ;-) I was more thinking about thinking about "will be back soon", "running to get your pdf", but a tipp of the day is nice idea too ;-)
@UlrikeFischer "If the pdf doesn't appear soon, blame Ulrike"
@DavidCarlisle I like the texts, but perhaps one can should add icons which are permanent over the languages.
1 hour later…
11:03 AM
@PauloCereda really it's a bit of mistery why they're called so (hispanoteca.eu/Foro/ARCHIVO-Foro/Palomitas%20de%20ma%C3%ADz.htm) but given that pigeons were raised with corn, and then cooked in oven... (evil, evil grin ;-P)
@Rmano oh no
11:39 AM
@CarLaTeX Thanks, I won't <3 Seeing the positive feedback here, I'll work on that as soon as I can :)
@PhelypeOleinik IMHO, an article like that is long overdue and it's great that you want to tackle that!
@Skillmonlikestopanswers.xyz :-)

I think that's the way the manual wants to see your code ( https://tex.stackexchange.com/a/552128/46023 ).

Can you check that out?
     /cis/.is family, /cis/.cd,
     domap/.is choice,
     /cis/domap/.is if=domap,   % new
%     domap/false/.code={\domapfalse},   % old
%     domap/true/.code={\domaptrue},   % old
     map/.default=true, % for this key
     domap=false, % global default
\newcommand{\cisset}[1]{\pgfkeys{/cis/.cd, #1}}

    \begingroup % change the keys locally
PS: All what I changed was
     /cis/domap/.is if=domap,   % new
%     domap/false/.code={\domapfalse},   % old
%     domap/true/.code={\domaptrue},   % old
12:10 PM
Probably yes...
@Rmano OK, thx.

If I choose: `\cmd[map=yes] `

I get (due to `/cis/domap/.is if=domap, % new` ):

`! Package pgfkeys Error: Boolean parameter of key '/cis/domap' must be 'true' o
r 'false', not 'yes'. I am going to ignore it.`

So I think my change was correct.
1:02 PM

But, if I repeat the stuff with a similar key it get Problems:
     /cis/.is family, /cis/.cd,
     domap/.is choice,
 %    domap/false/.code={\domapfalse},   % old
%     domap/true/.code={\domaptrue},   % old
    /cis/domap/.is if=domap,   % new
     map/.default=true, % for this key
     domap=false, % global default
% stroke order arrows
     dosoarrows/.is choice,
%    /cis/domap/.is if=dosoarrows,   % does not work
     dosoarrows/false/.code={\dosoarrowsfalse},   % works
     dosoarrows/true/.code={\dosoarrowstrue},   % works
Oh dear, what did I get involved with? :()

CommonStyle/.style={baseline=0em, anchor=base, remember picture},
\foreach \n in {\Startnumber,...,\Endnumber}{%
\node[] (tempcoordinate-\n) {S0(\n)};

\begin{tikzpicture}[CommonStyle, overlay]
1:23 PM
@PhelypeOleinik Very good! I'm looking forward to reading your article!
1 hour later…
2:49 PM
So what do you say now?

1. Can you think of the right content instead of S, x, y?

2. If yes: Do you have any ideas / suggestions as to which key would still be useful or needed here?

If interested, the code:
% arara: lualatex
% arara: lualatex
% arara: ghostscript: {  resolution: 250  , device: png16m}

% arara: cleanx: { extensions: [ aux , toc , log , bbl , bcf , blg , out , run.xml, gz, listing ] }


CommonStyle/.style={baseline=0em, anchor=base, remember picture},
% Default of Symbol/Map-Range
@Skillmonlikestopanswers.xyz -- Since I was one of those voting to close, I can tell you why I voted. The OP came up with a solution, and it wasn't obvious that a different answer would be likely. (It's likely I should have checked the time more carefully.)
3:05 PM
@PhelypeOleinik If tic-tac-toe is too long for the tutorial, maybe have a look at projecteuler.net for a couple of shorter examples? (I would love to see such an article!)
@samcarter_is_at_topanswers.xyz Oh, thanks, I'll definitely take a look at those! Chocolate covered candy seems a tasty one :-)
@samcarter_is_at_topanswers.xyz Your suggestion reminded me of rosettacode.org
@PhelypeOleinik Thanks for the link! This project looks interesting!
3:23 PM
@PhelypeOleinik The question looks completly useless ;-). Who cares about a coat of chocolate that it only 1mm thick and is calculated to the eight decimal?
@UlrikeFischer Definitely not enough chocolate :-)
3:53 PM
@UlrikeFischer The answer is simple: the values of a and b won't matter, one needs one complete bar in any case, because I'll eat the rest :)
@PhelypeOleinik ooh chocolate
4:11 PM
What have the Romans done for us? Apart from giving us page numbers: The introduction to T.L. Heath's The Works of Archimedes ends on page clxxxvi, before it switches to Arabic page numbers. It's the longest roman-numbered section I have ever seen in a book. Any better?
@HaraldHanche-Olsen Which definition of "better"? Longer?
@mickep I suppose … shorter is easy to find.
4:44 PM
@UlrikeFischer after \setmainfont[Script=Georgian,Renderer = Harfbuzz]{sylfaen.ttf} why do I get Missing character: There is no Გ (U+1C92) in font [sylfaen.ttf]:mode=harf;script=latn;language=dflt;+tlig;! with script=latn?
@DavidCarlisle you mean it is there with some other script? I need to check.
@UlrikeFischer no I don't know if it is the character is there but why didn't Script=Georgian at fontspec make `script=geor rather than script=latn at the font feature level?
@DavidCarlisle because the font doesn't have the script and so fontspec falls back.
Package fontspec Warning: Font "sylfaen" does not contain requested Script
(fontspec)                "Georgian".
5:13 PM
@UlrikeFischer ah right. there is so much font log I just searched for the missing character message, didn't look at the rest:-)
@UlrikeFischer so why does a font that lists Georgian specifically in its design aims not have that script tag:-)
@UlrikeFischer oh real life:( Georgian glyphs include khutsuri[citation needed] and mkhedruli letters. The development of the Georgian script came to a halt when Georgian type designer Anton Dumbadze was fatally struck by a car before being able to finish the design. As a result the font is incomplete.
1 hour later…
6:20 PM
@DavidCarlisle -- Sad story. You should add that to your answer about the font. It's relevant for future users who might try to use the font. (One article about Georgian scripts was published in TUGboat in 1998; I think there was another, but the title isn't obvious, and I don't remember the author's name.)
@barbarabeeton well, knowing the order in which who voted to close (and knowing of the comment and follow up question asked) I understand that you voted. Still I find the closing of a perfectly valid question strange (not only because it was about one of my packages, a fact I admit might influence my views on this), in fact, closing it as "solved elsewhere" or "solved by different approach" or something like that would have been perfectly understandable, just "solved as off-topic" seemed very weird
(earlier message was too long for a period, so here it is:) .
@HaraldHanche-Olsen that preface is longer than anything I ever wrote...
@DavidCarlisle in German a "Dummbatz" is an idiot. That type designer has a really funny name.
@Skillmonlikestopanswers.xyz -- The assignment to that category is almost always because that's the only one allowing a "free-form" reason. I don't like it either, but we're at the mercy of the site lords.
@barbarabeeton yep, the site lords who seem to always find a less pushing problem to solve than anything remotely useful.
@barbarabeeton but thank ye the lords that our site is now "accessible".
@Skillmonlikestopanswers.xyz -- I will be really happy when my new laptop is put into action and I can once again comment directly. (It arrived -- finally! -- two days ago, and my "sysadmin" husband is making sure everything is up to date.)
6:41 PM
@cis S equals kanas (S0 full symbol, S_i step by step), x equals arrows for the strokes of said kanas, y equals only the start point for such strokes. But honestly, my dayjob is enough guesswork, I don't need guesswork for my hobby as well. Are your strokes only available as TikZ code? I thought you created them in FontForge. If the former I guess sticking to pgf code is fine, else I'd like to avoid such a huge dependency and would rather stick to either LaTeX2e or L3 code.
@DavidCarlisle lesson 2 should be edited ...
@UlrikeFischer oh online systems yes will do, I did the help page but didn't notice that
@PauloCereda yes, much to the chagrin of my daughter Paloma
7:01 PM
@UlrikeFischer done
Sat Jul  4 19:02:53 UTC 2020
  Number of distinct users:     6
  Number of document requests:  61
  Number of documents returned: 27
  Number of unique user agents: 7
Fri Jul  3 23:47:07 UTC 2020
  Number of distinct users:     11
  Number of document requests:  54
  Number of documents returned: 46
  Number of unique user agents: 12
@UlrikeFischer ^^ you must be making a lot of tex errors (I'm sure I don't make any:-)
@Skillmonlikestopanswers.xyz Yes, everything is correct.

1) And I didn't want to pose a puzzle. I wanted to know if you have any suggestions for other useful or necessary keys. .... =>? .......
I could then add these.

2) Yes, I have everything as a ttf font (see picture).

If I had added this directly, the code would be much larger. So I wanted to leave it in general.

3) I use TikZ for the correct placement ("baseline, base, overlay").
And pgfplotstable for a list work (I'll show you later). But the commands are implemented separately. In itself, a TikZ free version can also be rewo
@DavidCarlisle I still wonder if we need a "hunt the error page", where people can learn to check for standard errors like ] instead of }, missing braces and similar.
@UlrikeFischer yes it would sort of fit lesson 2 but a bit mean to start with errors, unless you squeezed in a whole new lesson on erorrs and debugging (@JosephWright) ?
@UlrikeFischer one problem is that's something that is really better if you use a real editor, you can't easily even get line numbers on site (although I could probably fix that) so navigating back from an error message to the error in the source isn't as easy as it could be in other systems.
7:21 PM
@DavidCarlisle yes, but as your system shows the log-file, one could make some short exercises about some obvious errors like a typo or a missing brace. Perhaps with a screenshot how the error lines look.
@UlrikeFischer a new lesson 15, making the existing one 16 ?
@DavidCarlisle -- I'm pretty sure I suggested something about debugging a while ago, but got overridden on the grounds that it wouldn't be a good idea to discourage newbies.
@barbarabeeton possibly:-) but a lesson at the end probably wouldn't hurt, or a non-lesson in the extras bit by the gallery "gallery of common errors"
7:41 PM
@DavidCarlisle -- I agree entirely. The identification of a parenthesis instead of a brace is really high on the list (I find myself doing that distressingly often, and the font used in editor or browser on my screen isn't that easy to distinguish at a glance). Be sure to identify brace {}, bracket [], and parenthesis () -- non-native-English speakers often don't use the correct English terminology, but say just "brace" and expect it to be understood.
@DavidCarlisle -- Another must -- what is the anatomy of an error message. In other words, where does it begin, and what element in the message is the actual error. So many posts on tex.sx get that very wrong.
@barbarabeeton well I'm not so much about debugging more about "error happens don't panic".
@UlrikeFischer -- That's good advice. But encouragement to look for and at the log file can save a lot of tears later. Admittedly, I have a strong distaste for any "simple method" that tries to conceal the log file.
8:01 PM

Why does this not work?

\foreach[count=\N from 0]\n in {brown, red!70, orange, yellow!80!orange, green!70!black, blue!50, violet!60, gray}{\colorlet{strokecolor\N}{\n}}


\foreach[count=\N from 0]\n in {brown, red!70, orange, yellow!80!orange, green!70!black, blue!50, violet!60, gray}{\csname colorlet{strokecolor\csname N\endcsname}{\n}\endcsname}
@PhelypeOleinik \scantokens :) ...to all this, I must give a look to scontents, the LaTeX3 team works too fast...
@barbarabeeton as I'm sure @yo' will agree, good on line interfaces show the log on error, not hiding it behind an apparently working pdf file...
@DavidCarlisle @UlrikeFischer PR in ho-tex :)
@DavidCarlisle -- I haven't used such services, so I have no direct experience, only the plethora of questions here that look like the necessary information was never presented.
@barbarabeeton oh it was just a dig at @yo' as learnlatex.org just shows the log and no pdf on any error if you us (my) latexcgi button, but shows the pdf and a small error count in a red circle if you use overleaf. By the way does learnlatex.org work on your somewhat strange setup that doesn't like this site?
@PabloGonzálezL not sure I understood your comment "the changes only affect users who use perl pdfcrop and not those who use the direct version provided in TeXLive" isn't that the same thing?
@PabloGonzálezL also it would be helpful if you could say which issues this is solving what cases were wrong before that are right now, eg for cgwin it's clear the old code deliberately looked for a windows gs if a cygwin one wasn't installed and that's usually OK so long as you just reference local files not expect absolute paths, I think it will now say gs isn't installed in that case?
8:15 PM
Oops... my English is not as good as yours :(
I mean that the changes affect the direct use (from some windows terminal) of $perl pdfcrop.pl file.pdf
It doesn't affect anything if you use (from some terminal in windows) $pdfcrop file.pdf
@DavidCarlisle As it is, if you don't find GS in the windows PATH (under cygwin), try searching the registry using the WIN32 module, which is not available in cygwin since 2014
Also, if you capture gswin32c.exe for example, it doesn't work properly under cygwin (I'm not sure what's going on here, it seems to interpret /device as a path and not a gs option)
@DavidCarlisle -- I can read all the pages of learnlatex, but haven't tried out any of the suggested exercises. (But it's not only overleaf; there are a number of questions from people using lyx or scientific workplace.)
@DavidCarlisle That's why the comment only affects those of us who directly use perl pdfcrop, when running the windows encapsulated version (TeXLive) the proposed changes have no effect.
@barbarabeeton can you try to click on the latex.online button?
@PabloGonzálezL but it's better than @barbarabeeton's
8:31 PM
@UlrikeFischer -- I will try after we get back from dinner. (About to go out.)
@barbarabeeton try the button it's a new shiny version this week:-)
@DavidCarlisle -- Will do, but later, please. (I'm getting hungry.)
@PabloGonzálezL still not sure I understand, but will look later.
@barbarabeeton no rush. we are allowed to eat out from today as well:-) anyway aren't you supposed to be celebrating your freedom from colonial oppression?
@DavidCarlisle -- True, but this isn't the most cheerful July 4th I've ever experienced.

\foreach \N/\n in {0/brown, 1/red!70,
2/orange, 3/yellow!80!orange, 4/green!70!black,
5/blue!50, 6/violet!60, 7/gray}{
8:38 PM
@DavidCarlisle look what I found in pd1enc.def: \DeclareTextCommand{\textbeta}{PD1}{\ss} -- and this from a german ;-(
@UlrikeFischer well germans know that Greek is a one-letter alphabet
@DavidCarlisle ;-). We should get rid of the pd1 encoding and switch to unicode generally.
@UlrikeFischer yes
@DavidCarlisle :)
TeXLive 2020 and ghostscript installed (Win10 for example)
We clone the repository and want to play a little (just for fun :)

Scenario 1: cygwin without ghostscript in PATH and without having it installed in cygwin
pdfcrop tries to search the registry, fails to load `Win32::TieRegistry`
(not available for cygwin)

Scenario 2: cygwin with ghostscript in PATH and without having it installed in cygwin
pdfcrop tries to search the directories in the widndos PATH, finds gswin32c.exe
and it fails to execute (cygwin thinks `/`is a directory separator and not an option for ghostscript)
@DavidCarlisle That's more my reasoning... of course my mind works in Spanish :)
@MarcelKrüger I came across this issue today: github.com/MiKTeX/miktex/issues/577. It errors in texlive too. Do you know if lualibs has some means to extend the lua search path?
9:04 PM
@UlrikeFischer lualibs adds the function package.extraluapath(some_path) to add some_path to the Lua searchpath, but that does not interpret ?.
9:15 PM
@MarcelKrüger ah yes that works (once one doesn't forget to add the " " around the strings). I guess luapackageloader is simply incompatible?
@UlrikeFischer Not really, it's really the same issue as github.com/latex3/lualibs/issues/2
@PabloGonzálezL What do you mean? What changed?
@UlrikeFischer In contrast to the description of that issue, it is unrelated to package.path. In current LuaTeX versions, lualibs does not report missing modules properly because it still thinks that C modules can be loaded.
So whenever a name isn't found as a lua file, lualibs triggers a fatal error before luapackageloader has a chance to intervene.
@MarcelKrüger yes, I hadn't the impression that it is related to ?, I get errors also with simple pathes.
@PabloGonzálezL don't you think in English then google translate to Spanish? I thought that was the natural way?
9:26 PM
@DavidCarlisle Less so, but, in Spanish we have many synonyms and everything fits the context, I can't get used to the rules of writing English
@PhelypeOleinik For example we still have a {xparse}, but, much of it is already loaded into the latest version of the LaTeX kernel (am I right?)
@PabloGonzálezL In LaTeX2e, yes, but not on Plain or ConTeXt, so dropping xparse isn't an option (yet)
@PhelypeOleinik (yet) ...is part of the future plans? ...it would be great :)
@PabloGonzálezL Might be ;-)
@PhelypeOleinik Cool, updating today I saw a new package...it's the same one you worked on a while ago...I still don't read the documentation
@PabloGonzálezL You mean namedef? I had it written about a year ago, but didn't think it was worth uploading to CTAN... Seems like people liked it
9:40 PM
@PhelypeOleinik Hehehe...well now that namedef is in CTAN, you only need l3doc :)
@PabloGonzálezL Shush, don't need to remember everyone about that ;-)
@PhelypeOleinik Ops... I'm punishing myself :)
@PabloGonzálezL :-)
@UlrikeFischer I pushed a possible fix to the nocpackage branch (I'm not sure what the policy in lualibs is: Are the libraries patched or just imported?)
@MarcelKrüger we are simply importing them. The only thing that I still change manually is to reenable the unicode library (but probably this is no longer needed). So we probably should first try to get a change upstream.
10:23 PM
user image
I failed the comma :(
10:46 PM
@UlrikeFischer @barbarabeeton something like this? davidcarlisle.github.io/ll/en/extra-02
11:38 PM
@Skillmonlikestopanswers.xyz -- @DavidCarlisle will no doubt remind you. Otherwise, congrats.

« first day (3534 days earlier)      last day (35 days later) »